generator core repair alternatives 2

[Yuma]

Dear all,

We are in the process of repairing the core of a salient pole generator. This core has a bad hot spot that shorts some of the laminations. The shop offers two alternatives to do the repair:

option 1 - reuse the same core laminations, just relocating them in order to distribute the damaged plates all over the core and thus turn the hot spot into many little hot spots that hopefully will be negligible.

option 2 - build new laminations, dismantle the core until we reach the damaged zone, and replace the damaged zone.

I have never seen the case of option 1 myself, but I recall having heard of it. Does anybody have any experience of such a repair, and in that case, are you satisfied with the result?

As for option 2, the shop tells us that we could replace not only the damaged zone but all the dismantled zone - maybe half of the core- or even the full core with new laminations. Replacing the full core sounds like the best option technically, but my question is, given that the new laminations have 1 W/kg losses and the old laminations have 2 W/kg, do you think that, if we finally replace part of the core, we may have some issues like thermal or magnetical flux imbalance or whatever?

 

[ScottyUK]

I'm not an expert on salient pole machines but I would go with option 2 if this was a turbo machine. The majority of the costs are in dismantling the machine and stripping the core to a stage where it can be restacked; the cost of new core plates is a relatively small fraction. Knowingly introducing hotspots all over the core sounds like it's inviting trouble for the future in order to save a little now.

I wouldn't worry about the specific losses being lower, but do check that the magnetic properties are at least equal to those of the old core steel. I'd be surprised if they aren't. Also check that the losses are quoted at the same flux density.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Yuma]

Thanks, Scotty, I will check that with the supplier.

 

[electricpete]

The field seen by the poles is pretty much dc, isn't it?

I'm trying to think what scenario the laminations are important for... I'm sure the watts loss and hot-spot-performance during ac test are not as critical for this as they would be for ac core.





=====================================
(2B)+(2B)' ?

 

[electricpete]

I guess it is not the fundamental machine field of concern (which looks like dc to the rotor), but whatever slot ripples are present from passing the stator... as well as negative sequence from unbalance etc. These are generally relatively small.

=====================================
(2B)+(2B)' ?

 

[Yuma]

Hi Electricpete, thanks for your input

This is an ac core, i.e., I'm talking about the stator (armature) core. Maybe the way I explained it could have been a little confusing. When I said that this was a salient pole generator, it was just to give an idea of the dimensions of the core, whose diameter is much larger than its heigh, as opposed to a turbogenerator.

 

[electricpete]

Yes I read salient pole rotor and thought you were talking about rotor pole piece.

I don’t work in a repair shop... but we have sent our large motors to shop and had a pretty waide variety of core repairs done. I think the decision on each one is individual and we ask the shop to discuss the options considering both cost and long-term reliability.

Seems like there are some individual decisions:
1 - partial restack or complete restack. Depends on extent of hot spots. If cause is not known and suspect rest of core may not be far behind the hotspots, lean toward complete restack. If it is localized damage due to known cause such as electrical stator fault and rest of core looks great, only partial restack.

2 - new laminations or re-use laminations. New laminations tend to make economic sense for complete restack where whole core is in bad conditions. But ordering new laminations in small quantities for partial restack may not make as much sense. Laminations can be recoated when re-used if it is the interlaminar insulation suspect but that may or may not be the case.

If decide to resuse some laminations, would you put them all back in same place or shuffle thoughout (to extent does not require additional disassembly). Shuffling seems to make sense from one aspect: Once a hot-spot gets big enough, the temperature from hot-spot itself can be catalyst for further degradation especially C3 core plate. Even after repair of laminations they might still be suspect and spreading them out seems good.

There is a quote from Electrical Insulation for Rotating Machines that suggests some other reason for shuffling that I don’t quite understand:

Once the core is unstacked, the damaged laminations can either have their insulation restored or new laminations can be acquired to replace them. If the original laminations are reused, they are reorientated during the core rebuilding process so that the damaged sections are distributed throughout a large number of slots. This eliminates significant irregularities in the
slot or wedge grooves.

=====================================
(2B)+(2B)' ?

 

[rmw]

The shop may have suggested option because they knew that the original OEM of the generator didn't stamp those laminations any more. I ran into that with a company I used to work for. Most of the stamping die were either outside in the scrap heap or already being driven around as a new car in Japan.

There are companys that can laser cut new laminations reverse engineered, but they have to have a good lamination to use as a pattern. This means disassembling your machine and leaving it down while the new laminations are made. Normally electrical generation eqipment doesn't have the luxury of being allowed to remain idle for very long, hence that option.

I'm guessing of course, but been there done that, and it wasn't pretty.

As far as your second question, if you go with option 2, can you restack by equally distrubuting new and old laminations around the core? I wouldn't by any means put all new in one area and leave the old in another, but, I must say that when it comes to the electrical performance characteristics of core steel, I can hear the ice cracking around me.

rmw

 

[wolf39]

Yuma:

With 2 W/kg core iron material this generator is either very small in output or is very old (50 years plus). If the hot spot origin is punching burr you better don't distribute these faulty laminations to other locations. Instead you have to grind the laminations and to re-insulate them before reuse. It would help if you give us more information: Core diameter, core length, output, age, etc.

How did you locate the hotspot area? ELCID or high energy ring flux magnetization? What is a bad hot spot (hot spot temperature)? Sometimes hot spots can be "repaired" by splitting laminations with a knife and by inserting mica flakes there.

Wolf

http://www.hydropower-consult.com

 

[Yuma]

Thanks to all for your contribution, all your comments are greatly appreciated.
I think I should explain a bit more about the situation. This is a 6 MW old machine, and has gone through some incidents throughout its life, in fact it was working with two cutout coils, and when it had its last incident, we decided to do a rewind. At the time when all the old coils were removed, we found what I called a 'bad hotspot' which is a hole in the core, inside a slot. It is not caused by burr, but by a shortcircuit in a coil.
The hole is more or less the size of an egg, and affects two packets of laminations. All the laminations affected are welded in that zone. If the problem was in the surface of the core it would be easier to repair with mica flakes, as wolf39 says, but it is deep into a slot, and so unstacking the core is necessary.
So this is the situation. We are in the middle of the repair and the new coils are ready. We just have to make a decision, and it also depends on how many years this machine is planned to be working from now on, which I have to discuss with my bosses.

 

[edison123]

Unless you're willing to bone up for the cost of new core laminations + tooling cost for the same, restaggering is the way to go. I have done such repair/restaggering on small, medium and large machines, which are working well for years now without any problem.

Since the loss of new core is lower, you can work out the rate of return for such an investment. If it is attractive, then I would suggest full core replacement. (you can't go half the well in this case since the core loss and B-H will be different for each type of core).

Muthu

http://www.edison.co.in

 

[djs]

Bite my tongue, but I have seen a die grinder taken to the core and remove the welded area.

 

[electricpete]

“Electrical Insulation for Rotating Machines” discusses grinding as an appropriate localized repair alternative both in the slot as in original post and in the bore. Here is disussion of grinding in the slots:

High-Speed Grinding. High-speed grinding is effective for removing local core damage
in cores with open slots (if away from the wedge groove) and at the ends of cores with semienclosed
slots. The core surface shorts are ground away to provide separation between the
laminations. The ground area should then be treated with weeping epoxy and the depression
where the grinding was done filled in with thixotropic (high-viscosity) epoxy or other type of
filler to restore a smooth surface and prevent future winding mechanical damage.

Of course it is to be done only by someone that knows what the’re doing, and the post-repair core test is required proof of the repair.


=====================================
(2B)+(2B)' ?

 

[Yuma]

Grinding, as suggested by Djs and Electricpete, would be the option, if the damage was accessible. But it is inside a slot, and it is almost impossible to do it without unstacking the laminations.

 

[ScottyUK]

Hi edison,

What did you do to repair the core laminations prior to the re-stack - careful grinding / polishing to get back to a uniform thickness and true edge?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[lukin1977]

Can you send any pictures?

 

[TurbineGen]

Yuma - What are the dimensions of this slot? There should be many tools that could grind this out. I have done this on larger machines before.

Remember to treat the metal after you are done grinding. I'd use phospohric acid (you can run electricity through it to speed up the process). This will eliminate lamination shorting at the surface. If you are careful, you might hit it with muriatic acid first as this will accelerate the corrosion process (which is converted to an insulator by the phosphoric acid). After that you can use a weeping epoxy. This is the least expensive option. Much cheaper than a restack...

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[ScottyUK]

Twice in recent weeks I've heard the name 'muriatic acid' used in lieu of hydrochloric acid (a solution of HCl gas in water) on this forum. Is this just fashion, or am I so out of touch that the ancient names are returning to favour? Or do we have really members who studied chemistry in the 14th Century?

Just curious!


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[dpc]

It's actually a very commonly used term in the US anyway. I suspect you can buy "muriatic acid" in Home Depot stores anywhere in the US, but not "hydrochloric acid".



David Castor

http://www.cvoes.com

 

[ScottyUK]

Another example of us being divided by a common language! Thanks.


----------------------------------

If we learn from our mistakes I'm getting a great education!